The ability of tumor cells to invade and mestastasize is thought to be facilitated by proteinases that are either on the cell surface or are released into interstitial space. Intracellular levels of cathepsin B, a lysosomal cysteine proteinase, have been shown to correlate positively with metastatic potential in murine B16 melanoma variants as determined by their lung colonization potential. In these tumors the enzyme is abnormally increased in cell fractions containing plasma membrane derived vesicles. A variety of malignant human and murine tumors are also reported to secrete large amounts of latent and active forms of cathepsin B as compared with their nonmalignant and normal counterparts. When the intracellular and secreted forms of cathepsin B from tumor cells have been characterized, they are found to be significantly larger than the mature lysosomal form of the enzyme from normal tissues. These findings have suggested that many tumor cells have an abnormality in the post-translational processing of cathepsin B which leads to the miscompartmentalization and secretion of the precursor forms of the enzyme. Alternatively, cathepsin B in tumor cells may be a product of a different gene or may result from tumor specific differences in mRNA processing. In view of the probable importance of membrane associated and secreted forms of cathepsin B in tumor metastasis, I propose to compare the physical, chemical and immunological properties of cathepsin B in normal cells and in metastatic and nonmetastatic tumor cells; to compare the synthesis, processing and secretion of cathepsin B in these cells; to identify the defects which lead to the secretion of immature forms of cathepsin B by tumor cells; to correlate changes in intracellular sorting, processing and secretion of cathepsin B with metastasis in cells which express graded metastatic phenotypes; and to identify factors which regulate the secretion of cathepsin B. I also propose to compare the mRNAs and their in vitro translation products from normal and tumor cells in order to determine if they arise from different genes or from differences in mRNA processing. These results will provide unique information about the relationship between metastasis and cathepsin B processing and secretion, will be relevant to the secretion of other proteinases associated with metastatic tumor cells, and could lead to the improved clinical management of the metastatic cancer patient.